From Auto Racing To AI, These Four Leaders Bring Very Different Workloads To OCI

GoTo runs petabyte-scale streaming and lowers costs with flexible compute

Remote work pioneer GoTo migrated its global GoToMeeting video workload to OCI, which helped the company improve performance and cut costs by 50%. With nearly 800,000 businesses worldwide as customers, and tens of millions of active users, GoTo runs at petabyte scale on OCI, including multiple Kubernetes clusters across three cloud regions to provide low-latency access. OCI’s flexible, on-demand capabilities help GoTo maintain performance while reducing costs. “We were able to make use of Oracle’s flexible compute capabilities to match our workload needs precisely,” says Sascha Kuemmel, vice president of technology strategy. “Our workloads are very dynamic, due to the usage pattern of our customers. The reasonably priced on-demand model for compute in OCI is a perfect fit.”

OCI also helped GoTo save time and effort during the migration process. “We were able to simply extend our Kubernetes architecture on OCI without significant additional coding,” Kuemmel says. OCI’s network and security design, and its native Terraform automation, allow just two GoTo employees bring up entire regions in less than 24 hours. “The Oracle team has been a great partner in closing any remaining gaps and making the platform a perfect fit for our requirements.”

(Hear these industry leaders tell their stories on Oracle Live.)

Aleph Alpha builds self-learning AI using teraflops of training performance on demand

Headquartered in Heidelberg, Germany, Aleph Alpha researches, develops, and operationalizes large generalizable AI models that have the potential to solve humanity’s greatest challenges. To do that, the firm teamed with OCI to build its billion-parameter AI models.

Part of Aleph Alpha’s unique challenge was that it needed a high-security cloud provider, while also getting the best price-performance for intensive computing, says Jonas Andrulis, Aleph Alpha’s CEO and founder. OCI gives Aleph Alpha the freedom to cluster hundreds of bare metal CPUs and GPUs with low latency, creating a robust environment to pre-process and train its models. “OCI Compute provided the features we needed,” Andrulis says, including hundreds of GPUs with high-speed interconnect. “The ability to get over 140 teraflops of training performance on demand, and economically, has accelerated our work.”

Cox Automotive saves $4.5 million annually using automation and elastic scaling

Atlanta-based Cox Automotive implemented OCI to run several high-volume, back-office enterprise applications, including Oracle E-Business Suite on Oracle Exadata Cloud Service. This includes a real-time transaction engine that’s essentially the cash register for Manheim, the world’s largest wholesale vehicle marketplace, and a Cox Automotive business.

By consolidating and digitizing information related to vehicle transactions, Cox Automotive has changed the way the world buys, sells, owns, and uses vehicles. Known for its brands such as Autotrader, Kelley Blue Book, Manheim, and vAuto, Cox Automotive uses its unmatched data sets across the industry to provide insights and create services that connect dealers, lenders, and consumers in new ways.

Cox Automotive saved more than $4.5 million annually on back-office operations with OCI, in part due to the flexibility of the infrastructure, says Robert Taylor, lead back-office architect for Cox Automotive. For example, the company uses the powerful Oracle Exadata Cloud Service only during workday hours, turning it off on nights and weekends. Cox Automotive has been able to continually test automations, resulting in increased elasticity of its OCI infrastructure, such as provisioning 60 compute nodes to run for 2 hours at night and then shut down, or dial up the IOPS for Block Volumes only when needed instead of paying to provision them 24/7, as the company would in other clouds. “With OCI, we run workloads specific to the task at hand, save money, and gain reliability,” Taylor says. “Our customers can rely on our product, and we rely on OCI.”

Formula 1 racing team runs billions of simulations on OCI’s Arm-based virtual servers

World-champion Formula One race team Oracle Red Bull Racing uses OCI to analyze data for strategy decisions in real time and at scale. “For every lap and every corner that you turn, you’re always learning, capturing hundreds of data points a second,” says Ian Brunton, application development group leader for Oracle Red Bull Racing and Red Bull Technology. “We needed to make faster real-time decisions during races to get a competitive advantage, such as the best time to make a pit stop to change tires.”

To do this, Red Bull Racing ran its strategy simulations on OCI using Kubernetes and containers on Ampere Arm-based virtual servers. Using OCI let Red Bull Racing run about 25% more simulations during a race, giving strategists better insights to help them make the right call.

These simulations supported race strategists as they made decisions that ultimately helped achieve a winning performance in 2021, when drivers Max Verstappen and Sergio Perez collected 23 podium finishes, and Verstappen was crowned the F1 Drivers’ Champion. “With Oracle, we continue to be the most forward-thinking and innovative F1 team in the world,” says Brunton.

Join Oracle Vice President Leo Leung and these four industry leaders at Oracle Live to learn more about core infrastructure features that help them run applications faster, more securely, and with improved economics. You will also be able to dive deeper into compute, storage, and networking with presentations from OCI leaders and architects.